Building blocks for Network Privacy
summary
Surveys concrete network-privacy primitives, quantifies gains with k-anonymity and linkability metrics, and compares trade-offs in throughput and latency. Describes mixnets with exponential decay (example k≈100 batches, Poisson hold times) and 4KB/quarter‑second padding quantizations, shows MPC/DC‑net scaling to low hundreds of participants, and highlights homomorphic PIR progress: Spiral (Bitcoin/Ethereum balance queries) and Apple’s deployment for caller‑ID at 10–100 ms against multi-GB datasets. Notes practical costs: bandwidth blowup, committee trust, or 10–100× cryptographic overhead.
AI-written summary · about
chapters (9)
- 0:00Network privacy challenges
- 3:18Adversarial network model
- 8:17Privacy effectiveness metrics
- 12:12Mix networks evolution
- 15:42Padding to obscure sizes
- 18:57Multi-party communication
- 21:25Private retrieval with homomorphic encryption
- 24:43Recent advancements deployments
- 26:28Q&A
related talks
27mMetadata-private data transfer
08mPrivacy and Security
37mThe Thing with Privacy is
35mWhy Privacy Matters
47mPrivacy and Decentralisation with Multicast
30mBringing Privacy to Public Blockchain
54mNode Privacy, High-dimensional Graph Summaries, and Block Models
20mShare data without revealing personal information